As indicated in the prior section, there needs to be a reasonable amount of planning and design focus before a CDG process is ready to implement. Assuming that due diligence has occurred in the planning and design phases, let's now go through the key components of the implementation phase.
Process Readiness
Once again, there is nothing worse than a new governance council getting out the door only to be caught unprepared in a blind alley. In fact, it is very possible that once the CDG charter is communicated there will be an abundance of data quality and data management issues that emerge, which have been fostering for some time but have previously lacked an appropriate process into which they could be channeled. Be prepared to demonstrate the value and effectiveness of CDG on day one. To do this, the following readiness areas need to be addressed.
Communication of Charter and Implementation Plan
As with implementation of any major business process, there needs to be a communication plan to sufficiently broadcast the purpose and launch of the process. With the charter approved, it should be able to be internally posted and summarized for general communication purposes. Communication of the charter along with the expected implementation dates should come from the Customer MDM executive sponsor or sponsors. The executive sponsors, CDG Council members, and the MDM core team should all be clearly identified in the communication. The communication should also include or point to a clear description of how to engage the CDG process. It is important to emphasize any key scope and jurisdiction points in the body of the communication, so that the audience clearly understands the CDG purpose. For example, it may be important to emphasize that the CDG process is not a replacement for other existing IT and business decision-making bodies or steering committees.
Readiness of Process, Tools, and Baseline Measurements
Some process readiness checks may be necessary to fully validate how the process will be engaged, executed, and how various outcomes will likely occur. First, consider how issues and requests should be submitted to the CDG process. For example, what level of information and qualification will be required with the request? Can anyone fill out and submit a formal request or should a request be first reviewed with an organization's representative in the CDG council and only that representative can submit and sponsor the request? Either approach may be valid, but these types of process questions need to be considered and the expectations set correctly.
In any case, what's most important is that the CDG process will receive a well-qualified request that has business justification which can be acknowledged and further represented by a council member. Submitting issues that the council members cannot clearly understand, recognize, or consider sponsoring will only either demonstrate a disconnect between the council members and the requesting parties, or indicate a submission process problem where the detail and prequalifications needed to construct a solid request are lacking or not being enforced. Receiving well-articulated and qualified requests is a necessary first step to enable a good review and decision-making process. It's okay for the CDG process to reject poorly prepared requests, but always try to provide some feedback or further instructions that will help the requester understand the requirements for submitting a request.
Upon receiving a solid request, the CDG council should give the request appropriate review and quickly provide next step expectations to the requester. The request may require additional analysis or measurements to which often only certain data analysts, stewards, or IT roles can respond. This is another example of why an MDM core team and the regional data stewards all need to be part of the CDG network. These analyst and steward roles need to have sufficient tools and data access capability to conduct their assessments and provide accurate information related to the request. Chapter 6 will take a deeper dive into the actual process and tools used within a data quality management forum. That forum will typically be a sub-team function of the CDG process where many of the CDG requests will eventually go so that more specific analysis can be conducted or action plans initiated.
As the CDG process is ready to launch, be sure the necessary sub-teams, tools, data mining capabilities, and quality assessment measurements are ready to utilize as needed. Not being prepared with these capabilities could cause weeks if not months of delays while attempting to get other functions and resources to assist with these sizing and analysis needs.
Completion of Training and Readiness Plans with Core Teams and Sub-Teams
The council members, the MDM core team, and any related sub-teams or regional data stewards are likely to need some form of training or orientation to any new policies, standards, processes, tools, environments, documentation, metrics, metadata, or other collateral associated with the implementation and ongoing operational aspects of the CDG process. Ensure that all associated teams and key stakeholders are sufficiently familiar with the CDG related terminology, practices, and team members, and can easily get to CDG reference info.
Launch the Process
If all the prior mentioned elements related to planning, communication, and readiness have been sufficiently addressed, then the actual launching of the process should simply be a matter of having the executive sponsor send a final communication about the CDG process launch and generally thanking all sponsors, stakeholders, and team members who have contributed to the planning and implementation process. The various council members and any regional data steward teams can assist with any local channeling and launch communication needs.
Implement
Be sure that the implementation and ongoing process reflects good habits, maintains awareness of data quality trends and operational events that can impact or enhance the customer master data, continues to inform stakeholders of CDG actions and events, and keeps team members actively engaged so that the process of data governance is a continuously active and vital component of the overall Customer MDM practice. Next, we'll cover some of basic guidelines to follow.
Conduct Regular CDG Council Meetings
CDG should not be an as-needed process. In a Customer MDM environment, there are going to be more than enough challenges, need for policies, standards and quality improvement, ongoing data management and maintenance tasks, and so on, that should keep a CDG council, process, and associated teams continuously busy. Ensure that cross-functional and regional interests are being well served. Don't become focused on one region or one business area. There certainly will be issues and priorities that are associated with specific functions or regions, but in general, a Customer MDM initiative depends on the data governance process to drive global standards and consistent data management practices.
Over time, MDM and CDG influence will hopefully become well embedded in the company's business model with various processes and teams operating in self-monitoring and self-correcting modes. This will reflect that data governance has reached a mature and steady state. In this state the CDG council may not need to meet as frequently, but should always continue to keep a strong pulse going and clear focus on the Customer MDM environment.
Manage Priorities, New Issues, and Requirements
A robust CDG process should be regularly revisiting existing priorities and putting them into context with any new issues or requirements that have emerged. New needs come into play frequently that shouldn't be ignored or just added to the bottom of an existing queue. Expect that some new items may be of high priority and that some existing items may become less important or drop out of focus due to a change of need or circumstances since that item was initially submitted. In all cases, be sure to maintain an up to date list, ranking, and status of the items in the CDG queue, and regularly review this with the CDG council members and any other key stakeholders associated with the open items. In some cases, a large task or project in the queue may need to be broken out into smaller projects due to dependencies, resource constraints, or to better manage any associated impacts.
In general, when considering items for prioritization, make sure these fit within a reasonable time frame and expectation of execution. For example, for large data integration or data cleanup projects, there are often multiple phases associated with the overall execution plan. In such cases, it would be reasonable to focus CDG attention and priority just on the immediate needs and priorities associated with the current phase and perhaps the preparation needed for the next phase. Longer-term objectives often need more vetting out and are likely to be dependent on the execution and status of the current and nearer-term priorities.
Regularly Review Key Metrics and Performance Indicators
We previously mentioned that setting up key metrics and performance indicators is a necessary step in the CDG readiness phase in order to avoid the blind alley problem at implementation. And similar to having regular reviews of the CDG priorities and new requirements, the CDG council needs to keep a watchful eye on the key metrics and performance indicators from both a volume and quality perspective. In Chapters 6 and 8, we'll get into much more detail regarding quality management and the type of measurements needed, but it will be incumbent on the CDG council, and likely a data quality forum, to track the throughput and quality of the customer master data to fully understand and appreciate how people, process, and system events impact this data. And as is often said, it's very hard to control what can't be seen and measured.
We opened this chapter stating that you can't have an effective MDM practice without data governance. To add to that perspective, you really can't have effective data governance without having good measurements. So, be sure that key metrics and performance indicators are established and regularly reviewed in the CDG process.
Communicate Status of Projects and Improvements
Establish a broad and regular cadence of CDG communication to the sponsoring executives, core team, extended team members, and other stakeholders who will have interest or be impacted by projects and decisions. If anything, initially over-communicating about CDG is probably a good thing in order to raise awareness and to gauge who the interested audiences really are. Keep general communications succinct and well organized using a standard format as much as possible, and provide references or links to where more detail can be easily found if needed. Communications may need to be tailored for different audiences, but as a general rule don't inundate the audience with too much detail and be sure the audiences know who represents them on the CDG council or other sub-teams should they need to inquire about anything.
Communicate good and bad news quickly to ensure that awareness and opportunity for feedback is immediately available. Communicating about data quality improvements, project plans and achievements, new policies and standards, and so on, will generally be well received where the audiences have been advocating for and anticipating these results. But keep in mind that CDG actions or decisions can have a far-reaching effect and it's always possible that someone wasn't previously engaged sufficiently and may still have inputs or issues to raise that require a response.
Keep Sub-Teams and Regional Teams Actively Engaged
The CDG process needs to be active and participatory from the council members down to the front-end teams who enter the customer master data. That's not to say that CDG meetings should be conducted as large community gatherings allowing everyone from all levels to have their say. Rather, an enterprise-wide CDG process should consist of a top-level council supported by various interactive forums or sub-teams, which allow the extended teams and regional data stewards to participate effectively in the planning and execution of CDG initiatives. It's this active network of engaged data management–minded folks that creates the channeling and community framework that enables the whole MDM practice to thrive.
Because in Customer MDM there are typically many regional and local aspects to the data, processes, policies, and standards, there has to be a solid connection and level of interaction throughout the governance process with the regional and local teams.
Maintain and Improve
So far we have stressed the need to carefully plan, prepare, and implement a CDG model that will have sufficient recognition, reach, and impact across the enterprise. Achieving this objective is only half of the end state equation. Being able to also maintain a steady state, drive improvement in data quality, and increase data management effectiveness is the other half of the equation. Our next chapters will cover the details and techniques for maintaining and improving customer master data, but as with most everything that makes up and enables an MDM practice, it's the data governance team and process that has the responsibility to navigate this, make any necessary course corrections, and to ensure the MDM motor is well tuned and reaching its destination. Let's review some of the basics for maintaining and improving.
Complete Key Improvement Projects
While this may seem to be pretty obvious advice, in the data management arena it is not unusual for projects to be dropped or only partially addressed. As we have already mentioned, in the grand scheme of things data management and data quality improvement projects will typically take lower priority than other system, application, or operational projects. A key aspect of the CDG charter is to raise the visibility and level of importance for data management and data quality improvement projects. A well-established CDG council should be able to establish this, but will also need to demonstrate value by ensuring that these projects are well planned, funded, resourced, and executed in a timely manner. Successfully achieving this will depend on having those data analysts, data stewards, and sub-teams ready and able to be engaged.
Not having these resources and processes ready to be engaged will put the viability of the project into question or will create critical path issues that are likely to delay the project or reduce its scope. Failure of MDM projects to get initiated or sufficiently completed are signals of either an underlying issue with a company's commitment to MDM or ineffectiveness of the CDG council and process.
Success always breeds more opportunity, so be sure that the overall value proposition associated with the Customer MDM initiative and its data governance component can be successfully demonstrated through good planning and completion of projects that achieve the expected results.
Identify and Address Negative Quality Trends
Particularly in a large company, there is typically an expected amount of change or shift with customer-oriented operations. This can come from new systems and processes being introduced, organizational change, change in call center operations, change in partner models or vendors, mergers or acquisitions, or simply employee turnover. Any of these types of changes can impact what has been standard operating practice, which can have any number of impacts on the quality and integrity of the customer master data. If good data governance practices have already been well institutionalized, then many types of changes will still be subject to standards, training, and requirements that at least continue to maintain, and not degrade, the quality level of the customer data.
But where data governance is just getting off the ground and creating awareness, it will probably not have the proactive influence yet on many of these types of changes. This means that data governance will initially need to take more of a reactive stance when these changes occur and rely on its ability to monitor and respond accordingly if any negative data quality trends start to occur. This point further emphasizes the need to have good metrics, monitors, and regular reviews in place. Quick identification of a negative trend will allow the best opportunity to analyze the root cause and initiate corrective action plans.
Monitor and Correct Negative Data Entry Process Behavior
We talked earlier in this chapter about creating a data entry point matrix (see Table 4.1) that will indicate who, what, and where in regard to the process and authority to create, update, and delete specific data. This type of matrix in association to specific reports or monitors should serve as a guidance to determine if and where unexpected data entry behavior is occurring. In subsequent chapters, we'll provide more specific discussion and examples of this, but generally expect that either because of training issues or insufficient data access control, it will not be unusual to find authorized users who knowingly or unknowingly are not following data entry rules and standards.
Where the CDG council and the data stewards are already aware of where bad behavior is occurring, they should be prepared to address this quickly. However, being able to recognize these conditions and patterns is not often easy and may be a competency that develops over time. As long as there are sufficient baseline metrics, monitors, and reporting capabilities as part of the CDG process, the core team and extended data steward roles will begin to recognize bad behavior associated with the data entry points and practices. Recognizing this behavior is essential to developing control mechanisms that need to become part of an effective and mature CDG model.
Manage New Data Integration Requirements and Quality Impacts
Earlier in this chapter (Figure 4.2), we touched on how data governance can interact and serve a valuable role in the design and implementation of IT and business projects. And as we covered in Chapter 1, in an MDM initiative there is invariably some form of data integration that results in the creation of a master repository or system of reference. The CDG process should be highly engaged in the migration planning and integration process for any project that involves customer master data. Data integration efforts are one of the very best opportunities for data governance to become engaged and set the stage for data quality and integrity.
In Chapter 6, we discuss the dynamic between data governance with data quality management (Figure 6.4). This dynamic can be highly leveraged during large data migration and integration projects because these projects involve mapping of data from one source to another that will usually require business interpretation of the data and standards to be adopted where differences exist. A CDG council should ensure that any external data that will be integrated into the customer master data must meet certain data quality and integration standards. This is necessary so that the overall customer master data will not be significantly degraded by the integration of new data. Some quality and integrity compromises are likely to exist with integration of new data. The CDG council needs to be fully aware of this and engaged in decisions about such compromises.